Celeste: Variational inference for a generative model of astronomical images
Instrumentation and Methods for Astrophysics
2015-06-04 v1 Machine Learning
Abstract
We present a new, fully generative model of optical telescope image sets, along with a variational procedure for inference. Each pixel intensity is treated as a Poisson random variable, with a rate parameter dependent on latent properties of stars and galaxies. Key latent properties are themselves random, with scientific prior distributions constructed from large ancillary data sets. We check our approach on synthetic images. We also run it on images from a major sky survey, where it exceeds the performance of the current state-of-the-art method for locating celestial bodies and measuring their colors.
Cite
@article{arxiv.1506.01351,
title = {Celeste: Variational inference for a generative model of astronomical images},
author = {Jeffrey Regier and Andrew Miller and Jon McAuliffe and Ryan Adams and Matt Hoffman and Dustin Lang and David Schlegel and Prabhat},
journal= {arXiv preprint arXiv:1506.01351},
year = {2015}
}
Comments
in the Proceedings of the 32nd International Conference on Machine Learning (2015)